Research Methods

Question 1

What is a small effect size represented by for cohens d?

Answer 1

.2

Question 2

what is a medium effect size represented by for cohens d?

Answer 2

.5

Question 3

what is a large effect size represented by for choens d?

Answer 3

.8

Question 4

what is a small effect size for cohens r?

Answer 4

.1

Question 5

what is a medium effect size for cohens r?

Answer 5

.3

Question 6

what is a large effect size for cohens r?

Answer 6

.5

Question 7

what is a small effects size for cohens f^2?

Answer 7

.2

Question 8

what is a medium effect size for cohens f^2?

Answer 8

.15

Question 9

what is a large effect size for cohens f^2?

Answer 9

.35

Question 10

what do we mean by effect size?

Answer 10

a measure of magnitude of a result which is independent of the sample size

Question 11

What is it called if we accept the null hypothesis when we shouldn’t?

Answer 11

type II error

Question 12

what is it called did we accept a difference of relationship when there isn’t one?

Answer 12

type I error

Question 13

why cant statistical probability be used as a measure of magnitude?

Answer 13

it may reflect either the effect size of the sample size; meaning two identical studies have different significance due to sample size.

Question 14

why would we use a prospective power analysis?

Answer 14

to look ahead of what we need

Question 15

what is a sensitivity test also known as?

Answer 15

retrospective power analysis

Question 16

what does power depend on?

Answer 16

sample size, effect size and precision of measures

Question 17

what else do more reliable measures cause other than precise estimates?

Answer 17

smaller standard error

Question 18

what is standard error?

Answer 18

measure of the statistical accuracy of an estimate, equal to the standard deviation of the theoretical distribution of a large population of such estimates.

Question 19

Define the power of a test

Answer 19

the probability of avoiding making a type II error

Question 20

what power is desirable?

Answer 20

80%

Question 21

what power is desirable?

Answer 21

80%

Question 22

How would you know how many components/factors should be retained?

Answer 22

eigenvalues and scree plots

Question 23

What does a scree plot factors should be retained?

Answer 23

those above the debris

Question 24

what eigenvalues should be retained?

Answer 24

they should be above 1

Question 25

What do communalities measure?

Answer 25

the proportion of variance in an item that is accounted for by the selected factors

Question 26

When is a communality cause for concern?

Answer 26

if it is below .3

Question 27

Why would a psychologist not rotate initial factor loadings?

Answer 27

when the loadings represent a simple factor structure (no overlap in variance)

Question 28

In a results section for factor analysis what should you report?

Answer 28

amount of factors extracted and why, amount of variance it explains in % (eigenvalue =), inspection of scree plot, communalities and factor loadings which lead to a rotation or not

Question 29

what reasons could there be for non significant correlations?

Answer 29

wrong correlation test used, methodological reasons, test re-test reliability

Question 30

What tests can be used to examine two sets of scores?

Answer 30

t-tests

Question 31

what do t-tests do?

Answer 31

looks at the means for each participant to see if there is a significant result rather than correlation of all outputs.

Question 32

What is the purpose of a factor analysis?

Answer 32

to analyse patterns of correlations between variables in order to reduce these variables to a smaller set of underlying factors

Question 33

When would you use an explanatory factor analysis?

Answer 33

to identify smaller numbers of underlying factors when analysing a large number of items within a scale

Question 34

What is a communality?

Answer 34

each communality is the proportion of variance in an observed variable accounted for by selected factors

Question 35

what is an Eigenvalue?

Answer 35

sum of factor loadings within one factor for all variables (items)

Question 36

What is a factor loading?

Answer 36

proportion of variance in a variable accounted for by a factor

Question 37

what happens is as many factors are selected as variable?

Answer 37

each communality will be equal to one

Question 38

what is Kaisers criterion?

Answer 38

look at eigenvalues

Question 39

what is cattells scree test?

Answer 39

look at scree plot

Question 40

When do we need orthogonal rotation?

Answer 40

when there is a complex structure

Question 41

When do we need oblique rotation?

Answer 41

if correlations exceed .32 there is a 10% overlap in variance among factors so it is justifiable to rotate

Question 42

What does orthogonal rotation do?

Answer 42

redistributes variance among factors instead of first factor accounting for as much as possible, maximising variance with each factor

Question 43

Where can you find how much a factor explains variance?

Answer 43

in the 5 of variance column

Question 44

what is a non-monotopic relationship?

Answer 44

when it is like a curve or an arrow

Question 45

what is a monotopic relationship?

Answer 45

a linear relationship

Question 46

why would you use Spearman's rank?

Answer 46

because there is a relationship it just isn't linear

Question 47

What is the df for pearsons and spearmans rank?

Answer 47

N-2

Question 48

What would a pearsons correlation do if it had outliers?

Answer 48

overestimate the relationship

Question 49

what does a partial correlation do?

Answer 49

controls for a third value and measure the strength and direction between 2 whilst controlling for it

Question 50

what is the df for a partial correlation?

Answer 50

N-3